Making a Current Source Using an In-Amp

With the availability of many modern instrumentation amplifiers today, it has become quite an easy task to build a very precise current source. Current source is an important electronic building block that is supposed to deliver a constant defined current regardless of the load. If to a voltage source the low output impedance is the important characteristic, high output impedance would be for the current source. High output impedance means that as the load voltage changes due to current being sourced or sunk, there is very small change to the current. Ideally we would want the current to remain unchanged, because it is what it is supposed to do, to be a constant current source.

My Digilent Discovery Kit includes an AD8226 in-amp, and together with a buffer, OP02, also included in the kit, I made a very simple but very precise current source.

 What I like about this in-amp is it is a low-power, wide input and supply range instrumentation amplifier at +/-1.35 to +/18V, very flexible with the precision it offers. The Digilent kit has +/-5V fixed supplies that conveniently power the current source. The OP02 serves as a buffer to the reference pin of the AD8226. When you build your own current source circuit don't forget to always put bypass capacitors at the supplies.

               

Figure 1 Precision Current Source

                   

Here's how the circuit operates. The in-amp gives an output equals to the differential input multiplied by the gain, in this case I set it to one with pins 1 and 2 of the in-amp left open. The output of the in-amp is obtained across its output pin and the reference pin. The reference pin of the in-amp is buffered by OP02, and is equal to the voltage across the load. The OP02 is necessary in order to isolate the current coming out of the reference pin to the output circuit. It is important to note that there is a significant amount current flowing at the reference pin, and the output of OP02 provides a low impedance node. Providing a low impedance node means that even if there is a changing current at that node, the voltage at the node is unaffected and remains unchanged. In this case this voltage should solely be equal to the voltage at the top of the load. The output of the in-amp is then effectively impressed across Rs, which creates a current equals to Vin/Rs. This is the current that is passed to the load.

The circuit offers reliability and accuracy with the precision of the in-amp and the low temperature drift of the offset of the OP02. There are a few subtleties in the circuit one should be aware of. The programmable current is limited by the rails of the in-amp, the value of Rs, and the load resistance. The maximum programmable current will set maximum load voltage. This voltage, plus the voltage across Rs, should not exceed the rated swing of the in-amp. For example, if we have a 10k load, and an Rs = 10k, the maximum programmable current would be swing of the in-amp which is either positive or negative, Vmax, divided by Rload (10k) + Rs (10k). Another consideration will be the maximum power rating of the in-amp, which determines the maximum current that it can source. For the ad8226 it's rated at 140C junction temperature maximum, with 135C/W thermal resistance, allowing over 200mA of output current. 

Here's how the circuit looks like with my Digilent set-up.

In order to demonstrate the accuracy and linearity, at least generally and visually, I thought of applying a 0.4 Vp-p sinusoid signal at the input and observe the load voltage on an oscilloscope. The 0.4Vp-p corresponds to 40uA of peak-to-peak current, passed on to the 10k load which will produce the same voltage at the output. Using the oscilloscope software tool that also comes with Digilent, the output is shown in blue (C1) and the input in orange (C2).

Figure 2